Bioinspired poly(vinylidene fluoride) membranes with directional

Subscriber access provided by NEW MEXICO STATE UNIV

Interface-Rich Materials and Assemblies

Bioinspired poly(vinylidene fluoride) membranes with directional release of therapeutic essential oils Gianluca Balzamo, Helen Willcock, Junaid Ali, Elizabeth Ratcliffe, and Elisa Mele Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.8b01175 • Publication Date (Web): 29 Jun 2018 Downloaded from http://pubs.acs.org on July 5, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Bioinspired poly(vinylidene fluoride) membranes with directional release of therapeutic essential oils

G. Balzamo†, H. Willcock†, J. Ali‡, E. Ratcliffe‡ and E. Mele†*

†

Department of Materials, Loughborough University, Epinal Way, Loughborough, LE11

3TU, UK. ‡

Department of Chemical Engineering, Loughborough University, Epinal Way,

Loughborough, LE11 3TU, UK *E-mail address: [email protected]

1 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract Here, the morphology of Polypore fungi has inspired the fabrication of poly(vinylidene fluoride) (PVDF) membranes with dual porosity by non-solvent induced phase separation (NIPS). The fruiting body of such microorganisms is constituted of two distinct regions, finger- and sponge-like structures, which have been successfully mimicked by controlling the coagulation bath temperature during the NIPS process. The use of water at 10 °C as coagulant resulted in membranes with the highest finger-like/sponge-like ratio (53% of the total membrane thickness), whilst water at 90 °C allowed the formation of macrovoid-free membranes. The microchannels and the asymmetric porosity were used to enhance the oil sorption capacity of the PVDF membranes and to achieve directional release of therapeutic essential oils. These PVDF membranes with easily tuned asymmetric channel-like porosity and controlled pore size are ideal candidates for drug delivery applications.

Keywords: Phase separation; drug delivery devices; biomimetic structures; controlled porosity; plant extracts.


Page 2 of 36

Page 3 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Introduction The first time mankind employed fungi as source of food and therapeutic substances is dated back to primitive times.1-3 Since then these organisms have been thoroughly studied, especially for their bioactive molecules.4, 5 Many reports have revealed the beneficial properties of certain fungi for human health, including anti-tumour, antioxidant, anti-microbial, anti-inflammatory and immunostimulating activity.1-7 More recently, studies have focused also on analysing the morphology of fungi.8 The fruiting body of fungi basically consists of the spore-producing section (hymenophore) and the support sterile tissue (context).9 Within the group of Basidiomycetes fungi, the shape and morphology of hymenophore can vary greatly.10 In particular, the hymenophore of Polypores (Fig. 1a) possesses a well-structured tubelike morphology,11, 12 formed of cylindrical and uniformly-distributed capillaries that are supported by the fibrous context, as shown by scanning electron microscopy (SEM) images in Fig. 1b and 1c. This highly optimised architecture is capable of releasing around 30 billion spores per day,13 which is ten times higher than the amount released by fungi with a lamellar structure (e.g. Agaricomycetes release around 2.7 billion per day).8 Taking inspiration from the unique dual porosity of polypore fungi, here the fabrication of poly(vinylidene fluoride) (PVDF) membranes that are characterised by the coexistence of one region with long ordered micro-channels and one region with spherical pores is demonstrated. The biomimetic membranes were obtained by nonsolvent induced phase separation (NIPS),14-17 using PVDF in dimethylformamide (DMF) as the casting solution (CS) and water as the coagulation bath (CB). By controlling the thermodynamics of the phase separation process, the size of the pores was adjusted from 1 to 3 µm and the microchannel region was enlarged up to 53% of


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the total membrane thickness. Morphological and spectroscopic analyses were conducted to clarify the relationship between CB temperature and polymer precipitation rate. Although macrovoids have been already observed for PVDF membranes produced by NIPS,18-20 to the best of our knowledge a thorough study on the formation of a highly ordered array of microchannels distinctly separated from spherical pores has not been reported thus far. The presence of these well-defined and regular microchannels resulted in both an enhancement of the oil sorption capability of the membranes and the ability to directionally release antibacterial compounds. The bioinspired PVDF membranes with such dual porosity have the potential to be used as selective filtration devices and efficient drug delivery systems.

Experimental section Materials. PVDF (Mw=275000 g/mol by Gel Permeation Chromatography, average Mn=107000 g/mol, density=1.78 g/mL at 25 °C, melting temperature of 166-170 °C) and DMF (≥99.0%, density of 0.944 g/mL, boiling point of 153 °C) were purchased from Sigma-Aldrich and used as received without further purification. Water with analytical grade purity was purchased from Fisher. Galwick wetting liquid (surface tension of 15.9 mN/m) was supplied by Porous Materials Inc. (PMI). Highly refined mineral oil (Whitemor WOM 65) was purchased from Castrol Ltd, while tea tree oil (TTO) (100% pure) was purchased from Freshskin Beauty Ltd. Lysogeny broth (LB) and agar were purchased from Sigma-Aldrich. Escherichia coli B strain was purchased from NCBE, University of Reading (United Kingdom). Preparation of the bioinspired PVDF membranes. Casting solutions at polymer concentration of 19 wt% were prepared by dissolving PVDF in DMF for 1 hour under stirring (150 rpm). The PVDF solutions were subsequently deposited onto a stainless-


Page 4 of 36

Page 5 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

steel metal support by a 500 µm casting knife, and carefully immersed into a water bath. To prevent detachment of the membranes from the substrate and therefore ensure the formation of flat samples, loads were placed at the edges of the membranes when still in water. After 1 hour, the membranes were removed from the water bath and allowed to dry at room temperature under a fume-hood overnight. During the drying step, the wet membranes were kept attached to wet filter paper on both sides by means of pins, in order to reduce their shrinkage. The temperature of the coagulation bath (TCB) was varied from 10 to 90 °C in 10 °C increments; while the temperature of the casting solution (TCS) was 90 °C. All samples were stored in petri dishes at ambient conditions until characterisation. The samples produced are hereafter referred to as MTCS/TCB by considering the combinations of TCS[CB] used in the NIPS process. Membrane characterisation. The morphology of surface and cross section of the membranes was analysed by SEM (Hitachi TM3030, Tokyo, Japan) using an acceleration voltage of 15 kV, and a field emission gun SEM (Zeiss LEO1530VP) for the high magnification images. Prior to transfer the samples into the SEM microscope chamber, a 10nm-thick gold/palladium (Au/Pd) film was deposited (Quorum Q150T ES sputter) as conductive covering layer. ImageJ (National Institutes of Health, USA) was used to calculate the finger-like/sponge-like (FL/SL) ratio. The polymer precipitation rate, during the formation of the PVDF membranes, was measured by a Jenway 6705 UV-Vis Spectrophotometer. The light transmittance variation as a function of time was monitored through the nascent membrane submerged in water for 3 different values of TCB.


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Water contact angle (WCA) measurements were performed using a DataPhysics OCA 20 instrument with 1 µl droplet of de-ionized water. An average of 10 different zones per sample was analysed on both surfaces of the membranes. The membrane porosity () was measured by gravimetric method by estimating the weight of liquid contained into the pores using the following formula:21 ( )/

= (

)⁄ ⁄

× 100%

(1)

where is the weight of the wet membrane; is the weight of the dry membrane; !"#$%&

and

'()*

are the density of Galwick and PVDF, respectively. Three

measurements per each membrane type were performed using samples with an area of around 8 cm2. Galwick was used as wetting liquid due to its low surface tension (15.9 mN/m) and low vapour pressure (1-20 Pa).22 The oil sorption capacity (g/g) was investigated by measuring the weight of the wet membranes after 1 hour of immersion into oil. Prior to do so, the samples were left dripping for 1 minute to remove the excess of oil from the surfaces. The following formula was used for the calculation:23 +=

(, , ) ,

(2)

where - is the weight of the wet membrane after oil dripping and - is the weight of the dry membrane before oil soaking. The dynamics of oil absorption was analysed by light transmission tests. One side of the PVDF membrane (with sponge- or finger-like morphology) was exposed to white light, while one droplet of 1 µl of tea tree oil was deposited on it. The sorption process was then recorded from the opposite side by a HD camera. The oil infiltration into the membrane porosity determined a colour change of the membrane from dark grey to light grey. The brightness of membrane regions wetted by the oil increased as the amount of oil absorbed increased. The


Page 6 of 36

Page 7 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

frames recorded were processed by Photoshop software to determine the luminosity of those bright areas. The data obtained were normalised to the initial luminosity recorded. Release profile of TTO from the membranes. The release of tea tree essential oil from M90/40 membrane was determined using a UV-visible spectrophotometer (Spectronic Helios Alpha Beta UV-Visible Spectrophotometer, Thermo Electronic Corporation). A preliminary experiment revealed an absorbance peak at 215 nm of TTO which was used to detect the oil release from the membrane. A droplet of 4 µL of TTO was loaded onto the membrane from the sponge-like side. After the complete oil absorption, the membrane was let to float on DI water (water bath with a 10 mL volume). Aliquots were removed at multiple time points between 0 and 24 hours, and the oil concentration was evaluated through a calibration curve. Antibacterial tests. The antimicrobial properties of PVDF membranes loaded with tea tree essential oil were tested against Escherichia coli, as a model microorganism. Discs of 10 mm were cut from PVDF membranes and sterilised in a solution of 70% ethanol in DI water. A volume of 4 µL of TTO was loaded into each disc. Agar plates were previously prepared through inoculation of a single colony of E. coli from a fresh agar plate into 20 mL of LB broth. The bacteria were let to grow in a shake flask stirring at 225 rpm for 16 hours. Then, a volume of 100 µL of LB broth with 106 bacteria cells/mL was evenly spread out onto a freshly prepared agar plate using a L shaped spreader. The agar plates were dried out for 2 hours inside an incubator at 37°C. The PVDF/TTO discs were place in contact with the surface of the solidified agar medium. After 24-hour incubation, photos of the zone of inhibition were taken and used to evaluate the antibacterial activity of PVDF/TTO samples against E. coli. PVDF discs without TTO were used as control samples.


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Statistical analysis. For each set of results, independent trials were repeated at least three times. From raw data, mean and standard deviation (SD) were calculated. Oneway analysis of variance (ANOVA) followed by Tukey’s multiple comparison test with a confidence interval for mean of 95% were used for analysing the results by IBM SPSS Statistics software. The values of the membrane porosity are presented as mean ± SD of 3 independent trials. The values of the inhibition zone of the antibacterial tests represent means ± SD of 8 separate experiments. A significance level of p ˂ 0.05 was considered significant.

Results and discussion Effect of CB temperature on membrane morphology. The NIPS process used in this work to produce the biomimetic membranes is schematised in Fig. 2. The PVDF-DMF solution at 90 °C was first cast onto a flat metal substrate and then immersed in water to induce phase separation. The temperature of the water bath was varied from 90 °C to 10 °C and its effect on the morphology of the membranes was investigated. TCB variations are known to greatly affect the kinetics of the polymer precipitation process and they can be used to tune phase separation dynamics and consequently membrane structure.24, 25 As shown in Fig. 3a, when both the casting solution and the coagulation bath were kept at 90 °C (M90/90), the internal morphology of the membrane was dominated by closed spherical pores with an average diameter of 3 µm, forming a sponge-like structure, as reported in previous works.19, 26, 27 Macrovoid initiation started also to appear underneath the membrane surface in direct contact with water. By progressively decreasing TCB (M90/80 and M90/70 in Fig. 3b and 3c, respectively), the width of the sponge-like region diminished and macrovoids became more evident on one side of the membrane. When TCB was 60 °C (M90/60 in Fig. 3d), the macrovoids


Page 8 of 36

Page 9 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

were replaced by regular finger-like structures. Those features became more defined and narrow (average size in the range of 1-3 µm) by further reducing the temperature of the water bath (Fig. 3e-3i). As shown in Fig. 4, the width of the finger-like region increased by reducing TCB. The FL/SL ratio was negligible for M90/90 membranes (FL/SL≈2.5%), which were predominantly characterised by a sponge-like morphology. Instead, the FL/SL ratio significantly increased by reducing the CB temperature, until overstepping 50% for TCB≤20 °C. FL/SL ratios of 51.4% and 53.4% were measured for M90/20 and M90/10, respectively. Previous studies have shown the formation of macrovoids and fingers due to the NIPS process.19,

24, 28-31

In the work of Zhang et al., PVDF membranes prepared using a

casting temperature of 60 °C exhibited large and randomly distributed macrovoids, the cross-section of which was affected by the presence of water (non-solvent) in the casting solution.28 Jung et al. observed that macrovoids were formed for low PVDF concentrations (15-20 wt.%), whereas the addition of solvent into the coagulation bath determined the suppression of macrovoids.19 Our work highlights the possibility to accurately control the morphology and extension of regular finger-like structures by simply adjusting TCB. The emergence and preservation of narrow fingers with a precise size distribution, and the existence of a net FL/SL separation within the PVDF membranes have not previously been demonstrated. The orderly array of finger-like structures produced mimic, even if at a smaller scale, the array of microchannels that is characteristic of Polypores. In order to further investigate how polymer precipitation during the NIPS process affects the membrane morphology, light transmittance tests were performed.32, 33 Fig. 5 shows the temporal evolution of the phase separation during immersion of the


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

casting solution into the coagulant at various temperatures (inset of Fig. 5): TCB of 20 °C (blue symbols), 50 °C (orange symbols) and 80 °C (black symbols). Nearly 100% light transmittance was recorded at the beginning of the NIPS process for all three temperatures, due to the high transparency of the PVDF-DMF solution. When the phase separation started, the solidification of the solution resulted in an increased opacity of the cast material and, therefore, in a reduced light transmittance. The solidification of the PVDF-DMF solution began 15 s after immersion in water for TCB of 20 °C; whereas a shorter time (about 1-2 s) was measured for TCB of 50 °C and 80 °C. The polymer precipitation process lasted about 100 s, 55 s and 40 s for TCB of 20 °C, 50 °C and 80 °C, respectively. Therefore, the polymer precipitation rate increased by increasing the water temperature, in agreement with previous results.29 Polymer membranes produced by NIPS are characterised by the presence of a dense skin layer onto the surface in direct contact with the coagulation bath.34 The skin layer acts as a barrier to the solvent/non-solvent exchange and influences the morphology of the membrane.25, 35-37 The mass transfer associated to the in-flow of non-solvent into the CS and to the out-flow of solvent into the CB is mainly responsible for thermodynamic instabilities within the homogeneous polymer solution and for phase separation.38 As a result, solvent-rich and polymer-rich regions are formed within the phase-separating system, eventually becoming the voids (solvent-rich zones) and the solid matrix (polymer-rich zones) of the final membrane. In our case, at high CB temperature (80 °C), instantaneous liquid-liquid demixing happened, leading to initiation of macrovoids beneath the skin layer and formation of cellular morphology in the rest of the membrane (Fig. 3b).26, 37, 39 At low CB temperatures (20 and 50 °C), instead, physical gelation of the polymer solution occurred first in the NIPS process.29 This gave enough time to the initiated macrovoids to grow during the phase separation


Page 10 of 36

Page 11 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

process.40 The contraction force generated by the slow gelation of the surrounded polymer matrix compressed the macrovoids and determined the formation of fingerlike pores (Fig. 3e and 3h).26,

40, 41

By reducing the coagulation temperature, the

gelation process further impacted onto the membrane morphology.

Wetting properties of the membranes. Wettability and gravimetric tests were performed to investigate the effect of the diverse internal microstructures on surface properties, porosity and oil sorption of the bioinspired membranes produced. As expected due to the chemical nature of PVDF, all membranes were oleophilic, with oil contact angles lower that 10°. Indeed, when mineral and tea tree oil droplets were placed in contact with the membrane surfaces, they promptly spread over the surface and were absorbed. Water contact angle (WCA) measurements showed that the PVDF membranes were hydrophobic with WCA higher than 85° (Fig. 6a), as reported in previous studies.42 An additional distinction has been made between the wetting properties of finger-like or sponge-like side of the membranes. On the former surface, all membranes showed similar WCA values: 88°, 90°, 86° and 91° for M90/20, M90/40, M90/60 and M90/80, respectively. On the contrary, differences were recorded for the sponge-like side. While both M90/20 and M90/80 had similar WCA (around 90°), M90/40 and M90/60 were more hydrophobic (103° and 98° for M90/40 and M90/60, respectively). The increase in WCA can indicate a high surface roughness of the sponge-like side of M90/40 and M90/60, which in turn is expected to enhance oil sorption.43-47 Therefore, the oil sorption capacity of the PVDF membranes produced at TCB of 40 and 60°C was potentially affected by a balance between internal morphology and superficial characteristics. Both M90/40 and M90/60 membranes possessed higher porosity than M90/20 and M90/80 membranes. As shown in Fig. 6b,


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the porosity of M90/40 and M90/60 was (69.4±0.9)% and (68.5±0.7)%, respectively; whereas it was (59.8±1.3)% and (63.3±0.7)% for M90/20 and M90/80, respectively. The difference in porosity was statistically significant for M90/40 vs. M90/20 (p

Bioinspired poly(vinylidene fluoride) membranes with directional

Recommend Documents